I think one of the bigger problems will be the speaker rattle due to the mass moving.
maybe i can suggest a different arrangement which could work better tomorrow if i find the time
@Rennie alot of Manufacturers do things which are at least in theory suboptimal. The specific geneva design (the XL i suppose?) Might work with some algorithms to enhance stereo effects, but it does not seem like a design driven by pure technical facts rather than product design
maybe i can suggest a different arrangement which could work better tomorrow if i find the time
@Rennie alot of Manufacturers do things which are at least in theory suboptimal. The specific geneva design (the XL i suppose?) Might work with some algorithms to enhance stereo effects, but it does not seem like a design driven by pure technical facts rather than product design
Last edited:
@ saniii
I would love to meet up but I currently live in Austria far away from the german border.
The drivers arent "alone" above the tuning frequency, the impact from the PR (or port) just drops with 12db/octave.
So at 70Hz I get roughly 2,5mm excursion from the 6,5" PR = 32,2cm^3
and I get 3,8mm on the active 3" drivers = 36,5cm^3
For a total of 68,7cm^3 volume which again results in precisely 99db half space.
Thats the transfer function of the PR:
Anyways: even if I will just hit 94db (-6db like saniiis experience) it is good enough. Most off the shelf BT speakers dont go that low and not that loud.
The Anker Soundkore Boom2 for example (I got the opportunity to measure one) is tuned to 62Hz and only does 80db at that frequency.
Its close to 5L in volume and also weights 2Kg.
They use some Psy acoustic trickery but at anything more than 80db the bass is just lacking.
The slightly better JBL Xtreme 4 (same weight/size category) does around 85db better but still.
@ Rennie
I simulated all off dayton audios offerings and all are much worse in terms of MaxSPL or enclosure size needed.
The W3s are good, paired with the slightly bigger SB acoustics tweeter.
The distortion is much lower and i can get it flat to +-1db with just a few EQs.
The SB26STCN can go quite low at volume due to its bigger size and 1,2mm xmax. I currently cross steep (36db/oct) at 1Khz but could even go slightly higher if I get any heating issues but so far it looks very good!
Progress:
I tried the vacuum bagging on a sample part today for the housing. Its currently baking, I should be able to demold it tomorrow. Wish me luck, my first time carbon fiber.
Worst case if i cant get that to work is just doing a typical layup and painting the housing with some warnex or other structure paint. But i would really like the having the CF visible
I would love to meet up but I currently live in Austria far away from the german border.
Thats not false but there are some wrong assumptions:
The drivers arent "alone" above the tuning frequency, the impact from the PR (or port) just drops with 12db/octave.
So at 70Hz I get roughly 2,5mm excursion from the 6,5" PR = 32,2cm^3
and I get 3,8mm on the active 3" drivers = 36,5cm^3
For a total of 68,7cm^3 volume which again results in precisely 99db half space.
Thats the transfer function of the PR:
Anyways: even if I will just hit 94db (-6db like saniiis experience) it is good enough. Most off the shelf BT speakers dont go that low and not that loud.
The Anker Soundkore Boom2 for example (I got the opportunity to measure one) is tuned to 62Hz and only does 80db at that frequency.
Its close to 5L in volume and also weights 2Kg.
They use some Psy acoustic trickery but at anything more than 80db the bass is just lacking.
The slightly better JBL Xtreme 4 (same weight/size category) does around 85db better but still.
@ Rennie
I simulated all off dayton audios offerings and all are much worse in terms of MaxSPL or enclosure size needed.
The W3s are good, paired with the slightly bigger SB acoustics tweeter.
The distortion is much lower and i can get it flat to +-1db with just a few EQs.
The SB26STCN can go quite low at volume due to its bigger size and 1,2mm xmax. I currently cross steep (36db/oct) at 1Khz but could even go slightly higher if I get any heating issues but so far it looks very good!
Progress:
I tried the vacuum bagging on a sample part today for the housing. Its currently baking, I should be able to demold it tomorrow. Wish me luck, my first time carbon fiber.
Worst case if i cant get that to work is just doing a typical layup and painting the housing with some warnex or other structure paint. But i would really like the having the CF visible
You're assuming that the PR and driver outputs are in-phase and sum perfectly. Is that the case?
The carbon fibre + foam layering technique sounds like a great idea. Well done.
Chris
The carbon fibre + foam layering technique sounds like a great idea. Well done.
Chris
They are in phase above the tuning frequency (otherwise you would get a dip in the response).
The aren't in phase below tuning frequency -> the PR / port essentially act like a hole that freely lets sound/pressure escape -> acoustically "shorting" the front and back of the active driver -> that's why a PR/BR enclosure drop so fast below tuning frequency.
I couldn't wait and demolded my first part! Its not perfect but its only missing 10%. Its certainly usable and with a bit better technique i should get it looking even better.
I shifted the fabric while doing the tight corners (thats what i wanted to test) - I have a better feel for how much compound curve i can get away with so i will keep that in mind while designing the enclosure.
I did an "In mould Coating" process, the part comes out of the mould already finished with PUR paint to make it UV stable.
So you make the mould, add a release agent, paint the mould and than lay the carbon fibre.
I somehow ended up with airbubles between the PUR paint and the CF laminate. Looks like I didn't get enough resin below my first fabric layer. Its not super bad, essentially a cosmetic flaw but still bothers me.
I can tell you: the part is STRONG, it weights just 28g is made from 3 layers + the foam. The wall thickness is just 0,5mm on the sides and 3mm (due to the foam) on the bottom. It can easily support my weight.
I will make a second one to try and improve but it seems like i can make the housing like i want and it should be very strong.
The aren't in phase below tuning frequency -> the PR / port essentially act like a hole that freely lets sound/pressure escape -> acoustically "shorting" the front and back of the active driver -> that's why a PR/BR enclosure drop so fast below tuning frequency.
I couldn't wait and demolded my first part! Its not perfect but its only missing 10%. Its certainly usable and with a bit better technique i should get it looking even better.
I shifted the fabric while doing the tight corners (thats what i wanted to test) - I have a better feel for how much compound curve i can get away with so i will keep that in mind while designing the enclosure.
I did an "In mould Coating" process, the part comes out of the mould already finished with PUR paint to make it UV stable.
So you make the mould, add a release agent, paint the mould and than lay the carbon fibre.
I somehow ended up with airbubles between the PUR paint and the CF laminate. Looks like I didn't get enough resin below my first fabric layer. Its not super bad, essentially a cosmetic flaw but still bothers me.
I can tell you: the part is STRONG, it weights just 28g is made from 3 layers + the foam. The wall thickness is just 0,5mm on the sides and 3mm (due to the foam) on the bottom. It can easily support my weight.
I will make a second one to try and improve but it seems like i can make the housing like i want and it should be very strong.
Ah thats unfortunate...
The molding looks quite good, dont forget your respirator 😉
Do you plan do put in dampening material or something? Maybe some bracing around the sides?
The molding looks quite good, dont forget your respirator 😉
Do you plan do put in dampening material or something? Maybe some bracing around the sides?
So
I investigated the PR a little more and you guys might be right regarding "compression". However it looks like i have figured out why sanii experiences -6db and have hard data on it.
It seems like Dayton audio isnt really accurate with its data.
Lets take a look at the planed DSA175PR vs the Purifi PR.
The DSA has 8mm X-max while the Purifi PR got 15mm. I only need 8mm X-max so why bother with the expensive Purifi?
Sadly both manufacturers dont really explain how they rate/measure there X-max. On Active drivers its usually done by kipple measurements of suspension linearity and BL linearity. Usually the point where the stiffness/BL changes (what ever comes first) by 20% is characterized as X-max
I would assume this to be also true for PRs but as we will see its not.
But how can we get the data we need without buying a kipple test station?
Luckily both of these PRs have equivalent active drivers. And even better: both are measured!
Lets compare the DSA175PR and the DSA175-8.
The PR is rated at 8mm X-max while the woofer can only do 5,3mm, one could assume that the 5,3mm X-max is due to the BL changing not the suspension and therefore not applicable to the PR.
However i over on AudioXpress they measured the active driver and found:
Long story short: even with stretching the data for 20% THD, the woofer can only manage 4,1mm X-max before the SUSPENSION! linearity becomes limiting.
How does that translate to the PR?
It looks like the just pulled the 8mm X-max figure out of their butts... Or they assumed at 150% distortion or something wild like that.
At -7mm (the graph doesn't even go to 8mm) the suspension stiffness already increased from 0,6N/mm to 2,5N/mm which should be close to 3N/mm at 8mm.
So that would surely introduce issues!
Disclaimer: I assume that they just use the same parts for the PR and the active driver, they look very similar and are likely the same but i cant be 100% certain.
How does the Purifi compare?
Again we dont have data, they just claim 15mm X-max. The sister woofer PTT6.5W04-01A got measurements though!
The stiffness also changes quite a bit, the shape is similar but the curve is much less steep resulting in an increase of 0,4N/mm at from 0 to 8mm X-max.
This small change (compared to the ~2,4N/mm increase for the DSA175-PR) already indicated more than 10% distortion (cms is 66% down instead of 75% for 10% THD)!
So even the Purifi PR got some trickery going on, as it still changes a bit at 8mm and we can only estimate how bad it will be at the claimed "15mm X-max"
Erin over at Erinsaudiocorner.com also measured the Purifi woofer and the data seems to mostly align.
It tells the same story though. 15mm X-max is very much a strech, his graph goes to 13mm and shows an interpolated increase of ~2N/mm of stiffness.
Maybe we should use the woofer data instead of the PR data to simulate our enclosures/choose the PR to have better simulation/reality alignment.
At least for me it means i will shell out for the Purifi PR as it should be much better suited for my 8mm X-max application.
I dont like the fact that the membrane is paper, which would be very sensitiv to accidental mositure exposure. So i might coat the membrane with some PU paint or something to make it more resistant. Luckily the added weight of the paint isnt an issue as i can just remove the paint weigth from my tuning weight.
Maybe everyone is rating X-mech / X-Lim instead of true X-max for PRs!
So long
Farbe
I investigated the PR a little more and you guys might be right regarding "compression". However it looks like i have figured out why sanii experiences -6db and have hard data on it.
It seems like Dayton audio isnt really accurate with its data.
Lets take a look at the planed DSA175PR vs the Purifi PR.
The DSA has 8mm X-max while the Purifi PR got 15mm. I only need 8mm X-max so why bother with the expensive Purifi?
Sadly both manufacturers dont really explain how they rate/measure there X-max. On Active drivers its usually done by kipple measurements of suspension linearity and BL linearity. Usually the point where the stiffness/BL changes (what ever comes first) by 20% is characterized as X-max
I would assume this to be also true for PRs but as we will see its not.
But how can we get the data we need without buying a kipple test station?
Luckily both of these PRs have equivalent active drivers. And even better: both are measured!
Lets compare the DSA175PR and the DSA175-8.
The PR is rated at 8mm X-max while the woofer can only do 5,3mm, one could assume that the 5,3mm X-max is due to the BL changing not the suspension and therefore not applicable to the PR.
However i over on AudioXpress they measured the active driver and found:
Long story short: even with stretching the data for 20% THD, the woofer can only manage 4,1mm X-max before the SUSPENSION! linearity becomes limiting.
How does that translate to the PR?
It looks like the just pulled the 8mm X-max figure out of their butts... Or they assumed at 150% distortion or something wild like that.
At -7mm (the graph doesn't even go to 8mm) the suspension stiffness already increased from 0,6N/mm to 2,5N/mm which should be close to 3N/mm at 8mm.
So that would surely introduce issues!
Disclaimer: I assume that they just use the same parts for the PR and the active driver, they look very similar and are likely the same but i cant be 100% certain.
How does the Purifi compare?
Again we dont have data, they just claim 15mm X-max. The sister woofer PTT6.5W04-01A got measurements though!
The stiffness also changes quite a bit, the shape is similar but the curve is much less steep resulting in an increase of 0,4N/mm at from 0 to 8mm X-max.
This small change (compared to the ~2,4N/mm increase for the DSA175-PR) already indicated more than 10% distortion (cms is 66% down instead of 75% for 10% THD)!
So even the Purifi PR got some trickery going on, as it still changes a bit at 8mm and we can only estimate how bad it will be at the claimed "15mm X-max"
Erin over at Erinsaudiocorner.com also measured the Purifi woofer and the data seems to mostly align.
It tells the same story though. 15mm X-max is very much a strech, his graph goes to 13mm and shows an interpolated increase of ~2N/mm of stiffness.
Maybe we should use the woofer data instead of the PR data to simulate our enclosures/choose the PR to have better simulation/reality alignment.
At least for me it means i will shell out for the Purifi PR as it should be much better suited for my 8mm X-max application.
I dont like the fact that the membrane is paper, which would be very sensitiv to accidental mositure exposure. So i might coat the membrane with some PU paint or something to make it more resistant. Luckily the added weight of the paint isnt an issue as i can just remove the paint weigth from my tuning weight.
Maybe everyone is rating X-mech / X-Lim instead of true X-max for PRs!
So long
Farbe
Last edited:
There is an aluminium Purifi PR.
If you're going to coat things, then the Epique passive would be a choice.
Sometimes, "xmax" figure is the X mechanical limit of the suspension; after this and you'll be slapping the backplate or stretching the surround or crinkling cones.
Also that's one way x-mech ; needing 30mm total excursion on a 6.5 is huge lol.
If you're going to coat things, then the Epique passive would be a choice.
Sometimes, "xmax" figure is the X mechanical limit of the suspension; after this and you'll be slapping the backplate or stretching the surround or crinkling cones.
Also that's one way x-mech ; needing 30mm total excursion on a 6.5 is huge lol.
Usually Xmax is just a tad under Xmech for a PR.
A PR's distortion can't be measured without an active driver.
You can measure the excursion of the passive radiator, easy to do by eye, and measure distortion at different excursions.
Assuming a PR's increasing suspension stiffness causes massive distortion is not confirmed by actual measurements.
Never hurts to have too much excursion, when a PR reaches Xmech, it makes nasty "pup pup" sounds.
A port just makes more wind noise when over driven.
Yes. That said, some PRs are more linear than others.
We can look at some actual measured distortion data:
https://data-bass.com/systems/5c2e78ff80141f000411fc28
From what I've found, the TC Sounds 15" VMP passive radiator has a maximum peak to peak excursion of 89mm, an Xmax/Xmech of 44.5mm. Josh Ricci tested a pair driven by a single TC sounds LMSR-12".
The PRs were weighted for 18 Hz and 27Hz tuning (Fb), in two different tests.
His CEA2010 tests (SPL at maximum allowable distortion) at 2 meters would indicate the PRs were reaching ~37mm one way excursion, about 83% of their maximum around Fb.
The distortion around the 18Hz Fb was ~18.8%, with the 27Hz Fb only 12.5%.
As expected, distortion is higher at frequencies where the active driver's excursion increases.
The active driver's distortion is much higher than the passive radiator at Fb and the frequency range just above, below the active driver's maximum excursion.
In that case, no need to worry about the cheaper PR's distortion until it's hard limit of ~98 dB at 50Hz.
With added harmonics, should be able fake response down to 25Hz easily.
Art
The Al Cone Purifi PR is just 5mm x-max for 10% distortion while the paper cone does 6,9mm.
For 20% distortion we are looking at 9mm for the paper cone vs 7,6mm for the Al cone.
Thats based on Erins data, if you trust purifis data sheets, the difference is negligible. Not sure who to trust here.
The Epique PRs suspension is much too stiff and is low Q so not suitable. It also wouldn't need coating?
X-max should only be based on distortion.
X-mech/X-lim is the mechanical limit -> destruction beyond that point
Its also always single side.
So the 19mm X-mech on the Epique PR is really 38mm total travel.
At least dayton audio is calling it x-mech on that PR, Purifi is also calling it mechanical not linear X-max.
For 20% distortion we are looking at 9mm for the paper cone vs 7,6mm for the Al cone.
Thats based on Erins data, if you trust purifis data sheets, the difference is negligible. Not sure who to trust here.
The Epique PRs suspension is much too stiff and is low Q so not suitable. It also wouldn't need coating?
X-max should only be based on distortion.
X-mech/X-lim is the mechanical limit -> destruction beyond that point
Its also always single side.
So the 19mm X-mech on the Epique PR is really 38mm total travel.
At least dayton audio is calling it x-mech on that PR, Purifi is also calling it mechanical not linear X-max.
The SLAPS 8" PR has ~88 peak to peak excursion, ~44mm Xmax/Mech, more than double the Epique PR.
It just makes sense that it causes distortion. In a port, the air spring is perfectly linear. say 1N/mm no matter how far the active driver pushes the air. The restoring spring force stays the same.
However if the PRs spring stiffness changes, the pressure sine wave will be distorted. The peaks will be flattened if the spring needs more energy to push it farther to the peaks of the wave. Thats why we limit X-max in woofers not only by Bl changes but also by stiffness changes.
I calculated it before, that calculator just estimates the cone area based on a fixed factor.
The dayton audio should be good for what i want.
However after seeing what sanii said and reasoning why it might be true, it just makes sense to me that this change in suspension stiffness does a few things:
1. it adds distortion by changing how the resonator reproduces the sine wave, the peaks will get flatter because farther excursion requires more force than it should.
2. this increased stiffness also changes the tuning frequency based on excursion. A stiffer spring has a higher fs. So the change in tuning will also effect the system.
3. additional losses: if the tuning frequency doesnt change much/at all, the increased stiffness can only be the result of additional losses. These losses (lower Qmpr) will result in a SPL drop.
All of these combined can result in less restoring force than compression force -> the resonator will loose quite a bit of energy each cycle resulting in lower than predicted output.
Purifi also uses this as marketing. Sure sure they want to sell their stuff but still for me it makes sense that the system would act in that way with changing spring stiffness
Good thing is: the Purifi and the Dayton Audio PR have the same overall size.
So i will build my housing, test with the Dayton Audio PR and buy the Purifi PR if i dont get what the simulation says.
I will than measure and get true data on how this stiffness change changes the behaviour of the system.
To have low distortion you need 2x passive radiators, regardless of their fantasy measurements.
Dayton is more honest actually and poo ifi ramps up nonsense to ramp up the price.
At the end of the day, it is just a passive radiator, flapping away at 10 to 25% distortion.
It is why most " Hifi" builds stay away from them.
The point is not to sound like a Troll and just not waste time and over analyze passive radiators.
The solution is always the same 2x to even 3x more passive area to stay away from distortion.
From say 25 to 20% distortion to normal 3% to 10% you get from the actual speaker.
And with higher sensitivity speakers you can keep that down to 5% for high SPL
Xmax for a magnet transducer, is 10% at 70% magnet BL and you want to be under half of that, not at Xmax
for SPL goals.
The other solution for actual real bass in a small enclosure is massive magnets for .3 Qts
They can actually go in small enclosures. And not be confused by the quick roll off in sim.
The speaker Fs needs to be as low as possible, and with normal EQ the bass boost is actual real electrical/ mechanical reproduction.
Not flippy flappy fake passive bass, or poofy doofy fake port bass. People forget they are " assisted alignments"
Aka the bass knob getting turned up for bass boost. And the more bass boost that is real cone movement the better
No need to play with 3 woofers, but a good idea. Use one 8" woofer at 85 to 88 dB one watt not small 78 to 80 dB speakers
The Fs of a 8" is lower and any bass boost from EQ will be real. A compromise is possibly 7" drivers such as aluminum or wooven fiber daytons.
Again shooting for very low .3 Qts / massive magnets to work in a small box.
For clarity and image use 2x Dome Mids and 2x tweeters. Sensitivity will be way up there. Maybe 90 to 94 dB for a dome mid
higher of course for a tweet.
You either make the woofer louder with active power or make the mid tweet lower with typical padding.
With rear port, no passive, velocity can be high and less obvious rear ward.
It is good you know, woven fiber and epoxy construction. A slightly larger box and more sensitive speakers would be rather loud
with less needed power. Or enough power to match a 90 to 94 dB ish dome Mid which are screaming loud and clear with less than 4 watts
to a 85 to 88 dB woofer. Even higher to 92 to 94 dB 8" speakers but they need large boxes.
So you can stick to low .3 Qts speakers for smaller box and less SPL as a tradeoff
A dome mid needs a high crossover point, but much lower than a tweeter, so the "stereo image" is presented better with more mids
Then the woofer can be mono crossed around 500 to 800 Hz which is far far more stereo in the image than a 3k tweeter.
Vocals and instruments will be in actual stereo. With Dome mids and Dome tweets with very very high sensitivity
Dayton is more honest actually and poo ifi ramps up nonsense to ramp up the price.
At the end of the day, it is just a passive radiator, flapping away at 10 to 25% distortion.
It is why most " Hifi" builds stay away from them.
The point is not to sound like a Troll and just not waste time and over analyze passive radiators.
The solution is always the same 2x to even 3x more passive area to stay away from distortion.
From say 25 to 20% distortion to normal 3% to 10% you get from the actual speaker.
And with higher sensitivity speakers you can keep that down to 5% for high SPL
Xmax for a magnet transducer, is 10% at 70% magnet BL and you want to be under half of that, not at Xmax
for SPL goals.
The other solution for actual real bass in a small enclosure is massive magnets for .3 Qts
They can actually go in small enclosures. And not be confused by the quick roll off in sim.
The speaker Fs needs to be as low as possible, and with normal EQ the bass boost is actual real electrical/ mechanical reproduction.
Not flippy flappy fake passive bass, or poofy doofy fake port bass. People forget they are " assisted alignments"
Aka the bass knob getting turned up for bass boost. And the more bass boost that is real cone movement the better
No need to play with 3 woofers, but a good idea. Use one 8" woofer at 85 to 88 dB one watt not small 78 to 80 dB speakers
The Fs of a 8" is lower and any bass boost from EQ will be real. A compromise is possibly 7" drivers such as aluminum or wooven fiber daytons.
Again shooting for very low .3 Qts / massive magnets to work in a small box.
For clarity and image use 2x Dome Mids and 2x tweeters. Sensitivity will be way up there. Maybe 90 to 94 dB for a dome mid
higher of course for a tweet.
You either make the woofer louder with active power or make the mid tweet lower with typical padding.
With rear port, no passive, velocity can be high and less obvious rear ward.
It is good you know, woven fiber and epoxy construction. A slightly larger box and more sensitive speakers would be rather loud
with less needed power. Or enough power to match a 90 to 94 dB ish dome Mid which are screaming loud and clear with less than 4 watts
to a 85 to 88 dB woofer. Even higher to 92 to 94 dB 8" speakers but they need large boxes.
So you can stick to low .3 Qts speakers for smaller box and less SPL as a tradeoff
A dome mid needs a high crossover point, but much lower than a tweeter, so the "stereo image" is presented better with more mids
Then the woofer can be mono crossed around 500 to 800 Hz which is far far more stereo in the image than a 3k tweeter.
Vocals and instruments will be in actual stereo. With Dome mids and Dome tweets with very very high sensitivity
Last edited:
Id also recommend a two PR-Solution for the force cancellation to be honest. Something like the Dayton ND1xx or the Epique ones.
Regarding driver size itself, usually i end up with at least 4", mostly 5.5 or 6.5. If the space is there. Below, alot of Xmax is not really a thing and mechanical tolerances are much harder to handle.
But if the PRs are changeable, do it. I think there is no point atm to pay 100€ for a purify one, but the DSA175 is also not the one id go for 😀
Regarding driver size itself, usually i end up with at least 4", mostly 5.5 or 6.5. If the space is there. Below, alot of Xmax is not really a thing and mechanical tolerances are much harder to handle.
But if the PRs are changeable, do it. I think there is no point atm to pay 100€ for a purify one, but the DSA175 is also not the one id go for 😀
Except the DSA 175 driver is actually 88 dB not 79 dB like the current drivers, be louder with no combing.
And would make more decent bass with 38 Hz Fs
Far as I remember if you pair 1x passive DSA175 passive radiator with the 175 driver in correct volume.
The passive is already weighted correctly for a correct alignment. Have to check in sim
DSA175 driver is .29 Qts perfect example of a powerful magnet needed to use in a small enclosure.
As mentioned for low distortion, you actually need 2x passive radiator surface for every 1x woofer.
So that means 16mm Xmax not a single 8mm Xmax.
Which would be absolutely fine and be rather low distortion.
With 2x radiators you need to add and calculate the weight to get the correct alignment.
No different than any other 2x passive arrangement.
Being low Qts driver and assuming the usual 4th order alignment.
Tuning isn't rocket science it would be below driver Fs.
38 Hz driver be fine with 35 Hz tune likely.
Although of course a passive radiator would have a even sharper cutoff.
And would make more decent bass with 38 Hz Fs
Far as I remember if you pair 1x passive DSA175 passive radiator with the 175 driver in correct volume.
The passive is already weighted correctly for a correct alignment. Have to check in sim
DSA175 driver is .29 Qts perfect example of a powerful magnet needed to use in a small enclosure.
As mentioned for low distortion, you actually need 2x passive radiator surface for every 1x woofer.
So that means 16mm Xmax not a single 8mm Xmax.
Which would be absolutely fine and be rather low distortion.
With 2x radiators you need to add and calculate the weight to get the correct alignment.
No different than any other 2x passive arrangement.
Being low Qts driver and assuming the usual 4th order alignment.
Tuning isn't rocket science it would be below driver Fs.
38 Hz driver be fine with 35 Hz tune likely.
Although of course a passive radiator would have a even sharper cutoff.
Last edited:
@white, 88db is above a certain frequency. Often times the efficiency of Woofers is taken from >200hz above where the actual enclosure doesnt matter as much. if you take a look at the measurement graph, you can see very well at 50hz it is down to ~82db.
The needed vd for the passive radiators depends on tuning and electronic filtering. There is a certain tuning principle where the passive radiator is tuned a little lower than you actually would "from textbook" and EQing is used to push at the tuned frequency and use a sharp cutoff below.
For exmaple, this would be my current mobile speaker (Dayton Eqpiue 150HE with 3x SB16PFC PR) --> Alignment is "linear"´(given the room assumptions the tool is making). You can see a very balanced power curve with the peak occuring at f_tune. you can see this correlates well with the excursion minimum.
Max. SPL is predicted at 106db. Now... last but not least, the PR Excursion. At f_tune, it is well above its limit. with THREE of them. (3x*124cm² at 12mm vs. the active sub with 95cm² sd at 5mm excursion at f_tune).
you can go ahead, tune the system much lower to like 35hz where the uncorrected curve is more like a slow decline. now you add a high-q Filter with +6db at 35hz and a lowpass below that. what you will observe: 1. the actual amplifier power is now a lot higher. 2. the active speaker excursion doesnt change too much. 3. the PR excursion increases a lot. Overall, you are now limited by electrical power (both what your amp can provide and what the speaker can manage) as well as the PR Excursion.
So ... Taking 2x the Vd for the PR in comparision to the active chassis is a guideline, same as taking 1/3 sd for the reflex port. Its a starting point, but especially for more extreme systems or by using electronic filtering, 2x is not enough.
The needed vd for the passive radiators depends on tuning and electronic filtering. There is a certain tuning principle where the passive radiator is tuned a little lower than you actually would "from textbook" and EQing is used to push at the tuned frequency and use a sharp cutoff below.
For exmaple, this would be my current mobile speaker (Dayton Eqpiue 150HE with 3x SB16PFC PR) --> Alignment is "linear"´(given the room assumptions the tool is making). You can see a very balanced power curve with the peak occuring at f_tune. you can see this correlates well with the excursion minimum.
Max. SPL is predicted at 106db. Now... last but not least, the PR Excursion. At f_tune, it is well above its limit. with THREE of them. (3x*124cm² at 12mm vs. the active sub with 95cm² sd at 5mm excursion at f_tune).
you can go ahead, tune the system much lower to like 35hz where the uncorrected curve is more like a slow decline. now you add a high-q Filter with +6db at 35hz and a lowpass below that. what you will observe: 1. the actual amplifier power is now a lot higher. 2. the active speaker excursion doesnt change too much. 3. the PR excursion increases a lot. Overall, you are now limited by electrical power (both what your amp can provide and what the speaker can manage) as well as the PR Excursion.
So ... Taking 2x the Vd for the PR in comparision to the active chassis is a guideline, same as taking 1/3 sd for the reflex port. Its a starting point, but especially for more extreme systems or by using electronic filtering, 2x is not enough.
Both the Dayton ND and Epique ones arent suitable. Neither size wise nor compliance wise. They are just too stiff and lossy.
The force cancellation thing is what might be an issue. I will have to see.
Did you build a prototype so far? You also have issues with that as there is no way 3 PRs can be arranged to be force cancelling.
Doing to EQ trickery isnt doing much. Check out how the max SPL graph changes: it doesnt because you are still limited by physics.
Pushing more power isnt desirable on a project that aims to be light. You need more amping and more battery's to get to the same runtime, so more overall size and weight.
This is not my goal.
Same for the stuff WhiteDragon says.
Sure i can take a higher sensitivity driver. Heck lets go for a 2x 21" horn loaded tapped horn design for maximum efficiency, i would need just a few watts and no PR at all to get to 100db @ 50Hz, it even plays much deeper.
But guess what? Its at least 10 fold the size.
Even the DSA175-8 alone is 1/2 Kg heavier than the drivers i have chosen, thats without taking the bigger housing and 2x the PRs into account. I would also need much higher voltages to deliver 150W @ 8R instead of 3x 30W @ 4R -> more losses.
And what do i get? A single db more sensitivity compared to 3x 3" drivers with 50W more power handling -> really not worth it for my goals.
@saniiiii
Efficiency is a big deal for battery life.
Your "82" dB speaker needs 42 volts or 277 watts to reach 104 dB
And needs 14mm of cone travel to do that.
I recommended a 87 dB speaker for a reason.
To reach 104 dB that would be 70 watts
And needs a blistering 5mm of xmax to do it
Of course using max SPL in winisd is a fantasy regardless.
Nobody flattens their battery in 20 minutes with 70 or 277 watts.
Something really loud for a portable device would be 94 to 96 dB
That is about 10 to 12 watts for me and likely about 80 to 130 watts for you.
At 35 Hz my system is not unloading at 20 mm xmax with a abundance of power.
At 35 Hz xmax it is closer to 1mm of cone movement. Because that is what it is tuned too.
So when a typical " bass boost" of 3 to 6 dB EQ is added with a typical 40 Hz shelf filter.
My speaker does not unload at 30mm plus. It is = 2mm.
Hence when I say 2x passive radiators with 16mm of xmax is more than enough.
Heck actual 1 is enough for normal listening
It is because it is enough, at 10 watts or even 20 watts or 30 watts.
Few hours of battery turns into a few days, with such a massive battery to even fathom 277 watts.
Should we look at a larger box with a 92 dB speaker? maybe 6 watts is too powerful to maintain 104 dB much less than 277
Efficiency is a big deal for battery life.
Your "82" dB speaker needs 42 volts or 277 watts to reach 104 dB
And needs 14mm of cone travel to do that.
I recommended a 87 dB speaker for a reason.
To reach 104 dB that would be 70 watts
And needs a blistering 5mm of xmax to do it
Of course using max SPL in winisd is a fantasy regardless.
Nobody flattens their battery in 20 minutes with 70 or 277 watts.
Something really loud for a portable device would be 94 to 96 dB
That is about 10 to 12 watts for me and likely about 80 to 130 watts for you.
At 35 Hz my system is not unloading at 20 mm xmax with a abundance of power.
At 35 Hz xmax it is closer to 1mm of cone movement. Because that is what it is tuned too.
So when a typical " bass boost" of 3 to 6 dB EQ is added with a typical 40 Hz shelf filter.
My speaker does not unload at 30mm plus. It is = 2mm.
Hence when I say 2x passive radiators with 16mm of xmax is more than enough.
Heck actual 1 is enough for normal listening
It is because it is enough, at 10 watts or even 20 watts or 30 watts.
Few hours of battery turns into a few days, with such a massive battery to even fathom 277 watts.
Should we look at a larger box with a 92 dB speaker? maybe 6 watts is too powerful to maintain 104 dB much less than 277
I would have to look closer at the drivers, seemed to be around 79 to 80 dB at a quick glance.
So assumed no more than 84 dB with 3x drivers.
And whatever combing in the response with 3 radiators with different positions/phase cancellation
Seemed 6.5 inches is about the same size, so didnt want to push past a unreasonable size.
Just 87 dB compared to 84/85 dB
Actually real life below 250 Hz full space has 6 dB of losses.
So your typical WinIsd transfer function is just that. Transfer function of theoretical filter in Half space.
Forward only, no losses and diffraction losses, including baffle step is not seen.
Not real response. It would typically be a 6 dB bass boost in real life to match WinIsd.
But cone excursion and max spl can be calculated with boost to see how fast the excursion climbs
in sim.
Main thing was any noted issues with wanting more stereo image, which is why I added in using dome mids.
So stereo bandwidth includes more lower voice/ instruments. As with just tweeters and upper detail voice sibilance only
No need for guessing the 3 drivers and positions can be modeled in full space with actual frequency response data.
Not just transfer function for high pass filter alignment / vented enclosure
Last edited:
This is my PR vs port comparison, see link below.
PRs don't sudddently compress as heavily as ports do once the flow gets turbulent, but they distort with lots of harmonics at lower levels.
That's probably because of non-linearity of suspension.
Post in thread 'Investigating port resonance absorbers and port geometries'
PRs don't sudddently compress as heavily as ports do once the flow gets turbulent, but they distort with lots of harmonics at lower levels.
That's probably because of non-linearity of suspension.
Post in thread 'Investigating port resonance absorbers and port geometries'
Can I ask why you see the epique drivers as having poor compliance? I agree they are somewhat lossy (low mechanical Q), but the compliance value on the 5.5" version was tested via voice coil's test bench to have "XC @ 75% Cms minimum was 15.6mm" (direct quote), which seems more than adequate for the use case. If there's something I've missed in all this I apologise.